Positivity sets should have plotting files

Created by: Zaharid

Positivity files are weird. They are in the commondata format because people couldn't be bothered to have a specific format for them. They have no central value, no uncertainties and a sysfile with zeros. I'd be and not if they could be formatted and treated in a more sensible way. Which probably will have to wait until we nnfit.

The one think they don't have is PLOTTING files. As it turns out that would make sense. It would allow, well plotting them better. And also have some metadata that allows you to figure out what they are.

Another reason, relevant right now is that if we admit the existence of variants, so various alternative experimental values for a given dataset, then it makes sense to infer the identity of the data from the PLOTTING file, (which should have became META a long time ago).